The study intervention will be sited in three regions in Ghana: Eastern, Central and Greater-Accra regions. The intervention sites are public health facilities in the Lower Manya Krobo, Upper Manya Krobo district, and the Awutu-Senya East Municipality. The control sites will be in the Ada East, Shai-Osudoku and Effutu Districts. These districts have an average of 11% of their population being adolescent females 42 and reported adolescent pregnancy rates higher than the national average of 14% 22 . The intervention and control sites are selected health facilities which are matched on the type of facilities being district health centres and district hospitals. For each of the intervention sites, a senior member of the district quality of care team will be included as a resource person in the RMC trainings.
The study is a non-equivalence pre-test post-test quasi-experimental study. Both qualitative and quantitative methods will be used in data collection.
Quantitative study population
Study participants in the quantitative phase will be adolescent females between the ages of 10 and 19 years old who deliver at the participating facilities.
Qualitative study population
The first group of participants in the qualitative phase of this study are adolescent females between the ages of 10 and 19 years old who deliver at the participating study sites. The second group of participants are health providers in full time employment at the study sites. Adolescents and health professionals alike will be invited to participate in the study based on the study inclusion criteria and will be enrolled based on their consent to participate. Adolescents will be invited to participate in In-Depth Interviews (IDIs) and the professionals will participate in Key Informant Interviews (KIIs).
Adolescents will be deemed eligible for participation if they satisfy the following criteria
- Between the ages of 10 and 19 years
- Present at any of the selected health facilities for delivery, antenatal care within their third trimester or post-natal care within two weeks of delivery
- Consent to participate in the study
Eligibility criteria for health professionals are
- Health providers including midwives and administrators who are in full time employment at the selected health facilities
- Consent to participate in the study
Health professionals and adolescents will be excluded based on the following criteria
- Decline to participate
- Are physically or mentally ill or indisposed
- Did not deliver at the facilities in the study if they are adolescents or are not obstetric care providers or administrative staff of the facilities
Sample size determination
The minimum sample size for adolescents is obtained observing the assumptions that the estimated proportion of women who encounter mistreatment during childbirth is 18%43 and the proposed intervention may reduce this by 10 percentage points. Additionally, a 5% margin of error, a 95% significance level and 80% power are estimated.
N= estimated sample size for one arm
Za=Critical value at confidence interval of 95% (1.96)
Zb=Critical value at power of 80% (0.842)
d= Actual difference in proportion between control and intervention groups (0.1)
P= Average of difference in proportion of both groups
An intra-cluster design effect of 1.11 and a non-response adjustment of 10% is factored to obtain a minimum sample size of 392.
A minimum of 10 healthcare participants and 10 adolescents will be included in qualitative Key Informant Interviews It is anticipated that the number of interviews will allow a sample size large enough to allow data saturation to understand the phenomenon of mistreatment from both provider and client perspectives.
Multistage, stratified and systematic sampling will be used to recruit study participants from the facilities where they attend. Using data available from the 2010 Population and Housing Census, three out of the formerly ten regions in Ghana with adolescent pregnancy rates higher than the national average (14%) were randomly selected. Subsequently, six districts were purposefully selected from these regions considering appropriateness for the intervention, available resources, and geographical non-proximity between intervention and control facilities. Facilities in the intervention and control arms of the study were matched based on the number of adolescent pregnancies delivered annually using the Ghana Health Service District Health Information Management System (GHS-DHIMS) data from 2018 as a guide.
All eligible adolescent participants who attend the participating intervention and control facilities for antenatal care in their third trimester and delivery will be invited to enrol in the study by a trained research assistant. The nature of the study and all ethical concerns will be explained to the candidate prior to their enrolment. Enrolment will be run on a cumulative bases until the minimum sample size is satisfied. Some adolescents will be purposively selected to participate in the adolescent IDIs based on their characteristics which is aimed at achieving maximum variability of participants. Each district will have a health professional who works in administrative capacity and in a caregiving capacity invited to participate in the Key Informant Interviews.
Quantitative data will be collected with the aid of an electronically-based structured questionnaire adapted from Bohren and colleagues’ study 44 which has been modified to suit the local context. The tool is designed to collect data on the socio-demographic characteristics of adolescents and their knowledge on health and sexual reproductive rights. The tool will additionally collect data on the labour experience including the provider-patient encounter and relationships, encounters of mistreatment and RMC, and health facility factors.
Qualitative data collection will be done with the aid of interview guides adapted from the studies by Bohren et al 44 and Afulani and colleagues 45. The tool is specifically developed to collect data on the varying forms of mistreatment within the local context. Sections of the guide probe for an in-depth discussion on the themes of knowledge on health and sexual reproductive rights, provider-patient interactions, encounters of mistreatment and RMC and health facility factors. For professionals, the interview will probe for the role of healthcare providers in the labour experience especially in the administration of RMC.
Data collection procedure
The quantitative interviews will be administered by trained research assistants. Interviews will be administered both at baseline and endline (from day 1 following the intervention). They are estimated to last between 30 and 45 minutes per participant and will be held at the home of the participant or a convenient private venue outside the health facility within two weeks after delivery. This venue arrangement is considered important to create an enabling environment which allows free and candid dialogue with the participant which may not be possible at the health facility due to fear of retaliation from providers. This time interval between delivery and the interview is also considered adequate for the parturient to recuperate without introducing significant recall bias. Interviews with adolescent participants will be in the most commonly spoken languages in the district which are English, Twi, Dangme or Krobo, based on the preference of the participant. This will allow the participants to express themselves freely in the language they are most comfortable with.
Qualitative IDIs will be conducted with selected adolescent participants and KIIs with health professionals. Interviews will be audio-taped, with prior permission from the participant. Additionally, notes will be taken to capture non-verbal nuances and expressions which may aid in better analysis and interpretation of the audio information. The number of interviews will continue until a saturation of information is reached. Each interview is estimated to last between 45 and 60 minutes. For adolescents, IDIs will take place at a location convenient to the participant outside the facility within two weeks after delivery. Similar to the quantitative interviews, the qualitative interviews will be conducted within a post-recovery timeframe which mitigates the probability of recall bias. Based on the preference of the participant, IDIs will be conducted in English, Twi, Ga-Dangme or Krobo as done in the quantitative interviews. The KIIs will be conducted within the facilities were the staff work and will be conducted in English.
Data quality assurance activities
The quality of the study and data collected will be assured via the following means
- Intervention-Control Allocation: The intervention and control facilities are selected based on the similarity of their characteristics. This includes the type of facility, proportion of adolescent deliveries per year and the socio-demographic characteristics of the community. The geographic non-proximity of the facilities was also highly considered to avoid contamination of the control communities.
- Research Assistants Recruitment, Training & Monitoring: Research assistants will be selected based on the following criteria
- Sex: female
- Age: 18 to 25 years
- Academic qualification: minimum of a secondary school certificate
- Technical proficiency: smartphone literacy
- Language proficiency: at least one of the languages (Ga-Dangme/ Krobo/Twi)
The age group and sex of the assistants will help establish rapport between the assistants and participants. This will also help minimize response desirability bias on the side of participants. The educational and technical criterion are necessary to ensure that the assistants understand and effectively carry out their roles in the study to collect quality data. The assistants will undergo a two-day training on the theoretical background of the study, enrolment and obtaining consent from participants, administration of questionnaires and facilitation of interviews, as well as ethical issues. The principal investigator will always closely supervise the research assistants to ensure their observance of study protocols and ethical codes
- Data Collection Instruments: The data collection instruments are adapted from a previous multi-country study (which included Ghana) which included parturient women of all ages 46 instruments will be adapted for use with adolescents. The tools will be pretested on 5% of the sample size in a pilot facility in the Kpone-Katamanso district and modified accordingly based on findings from the pilot. The pilot district was selected based on its similarity to study sites in having an adolescent birth rate of greater than 11%.
- Data Collection Timing and Location: Data collection at baseline will be over a 3-month period. Endline data collection will start from day 1 after the intervention until three-months post-intervention. The venue and time interval between delivery and the adolescent surveys will allow participants express themselves freely without fear of victimization by health providers. This will also allow the participants enough recovery time from childbirth with minimal recall bias.
- Data Monitoring: Data collection will be conducted in real time electronically. The data will be monitored daily to allow early detection of discrepancies and resultant remedial action. Transcripts from audio interviews will be reviewed by an independent listener who is fluent in the language and did not conduct or transcribe the interview to audit for accuracy and completeness.
Data processing and statistical analysis
Both baseline and endline data will be reviewed and edited for completeness and consistency daily. The primary outcome variable of interest is defined as disrespectful intrapartum care which is measured as a response of “Yes” to any of the nine mistreatment questions. The explanatory variables of interest include socio-economic status, educational level, marital status and knowledge of health rights. Data will be collected on the KoBoCollect platform, synced onto a web server and downloaded in Ms-Excel format to the principal investigator’s computer. The data will be exported in STATA version 16 for cleaning and analyses. Descriptive statistics as well as binary and multiple logistic regressions will be computed. Additionally, Difference in Differences (DID) will be calculated to determine the net RMC training intervention effect on mistreatment and RMC. Finally, statistical significance of the RMC training intervention effect will be evaluated using adjusted prevalence ratio, adjusted odds/risk ratio at 95% confidence interval (CI) and P-Value of less than 0.05.
Audiotapes from qualitative interviews will be translated and transcribed verbatim into Ms-Word format in English. Themes will be identified and coded in Ms-Excel and exported into NVivo for analyses. Quotes will be narrated and cited in participant codes.
Data analysis for qualitative and quantitative data will be conducted separately. Subsequently, inferences from both qualitative and quantitative results will be jointly used to draw conclusions.
The proposed intervention aims to educate and inform midwives about the concepts of mistreatment and RMC. The training additionally seeks to help participants explore and shift their current values and beliefs about mistreatment of adolescent parturients towards more positive behaviours. It also aims to encourage the practice of RMC by promoting adherence to positive professional codes of conduct, ethics, roles, values and attitudes. The proposed intervention is a training module facilitated with the aid of a resource package adapted from the White Ribbon Alliance/ USAID “Respectful Maternity Care for Healthcare Workers: Tackling Disrespect & Abuse During Facility-Based Childbirth” project 47. Additionally, the structure and content of the module is influenced by evidence from previous studies and systematic reviews on mistreatment and RMC 48–54. This course will be presented in the format of presentations, case studies, role plays, discussions and evaluations. The content will cover the adolescent and their reproductive health rights, the Ghana Health Service (GHS) code of professional ethics, Nurses and Midwives Code of Conduct, definitions and typologies of mistreatment and RMC as well as their outcomes to the woman, neonate, professional and health system. The content will also cover the professionals’ role in the delivery of RMC. The training workshops will be implemented by the principal investigator supported by a facilitator who is member of the District Quality of Care team. Each workshop will be a two-day event. The first day will cover a theoretical introduction of the concepts of mistreatment and RMC and the role of the obstetric professional. The second day will be a series of case studies and role plays to apply the theoretical basis from day 1. The facilities will additionally be provided with visible posters and materials as a constant reminder of the contents of the training programme.
There will be no intervention conducted in the control facilities.
It is anticipated that at the end of this proposed study, the following outcomes will be achieved per each objective as summarised in Table 1
Insert Table 1 here